During the first five years of this project, we compared the efficacy of a targeted, early intervention strategy using neuroleptic medication with a strategy using early intervention (increased dosage) that supplemented ongoing maintenance neuroleptic therapy. Our experience to date is useful for the work proposed in the current application in several ways. First, we have developed a rapid, effective pharmacologic intervention at relapse for patients assigned to either drug-free or experimental drug status. Since targeted, early intervention treatment efforts are feasible and effective, the consequences of ineffective experimental or placebo treatment need not be severe deterioration in clinical state. Second, our experience with the early intervention, targeted neuroleptic strategy suggests the likelihood that highly sensitive detection of a prodromal syndrome includes symptoms that are false-positive with respect to relapse and often includes psychopathology likely to be susceptible to anxiolytic drug treatment. Third, we now have a trained and experienced clinical staff with established competence in dealing with chronically and severely mentally ill patients within the protocol demands of controlled clinical studies. Fourth, we have gained the logistical experience necessary for conducting outpatient clinical trials. As originally described, the specific aims of the project were to determine the relative effectiveness of two pharmacotherapeutic approaches to the care of schizophrenic outpatients, one involving the targeted use of neuroleptic drugs for symptom exacerbation and the other involving the use of these same drugs for maintenance purposes. At the outset, the feasibility and ethics of the noncontinuous targeted approach were heatedly debated; however, the potential advantages of this alternative strategy for the actively noncompliant patient were compelling, and there was a growing appreciation that placebo controlled continuous drug trials had not addressed the central hypothesis of this project. The possibility that this approach would be a feasible alternative for a substantial proportion of schizophrenic patients was sufficiently attractive to merit empirical evaluation in a random assignment, comparative treatment study. The immediate goal was to ascertain if a substantial reduction in the use of neuroleptic drugs could be achieved without adversely affecting course and outcome status as reflected in symptomatic clinical decompensation and the need for hospital care.